precision shaft engineering

colin · 2007-03-08T22:17:53+00:00

Hi,Im realy an electronics enigineer but my project has a shaft wich is 1 meter long 5mm dia and running at steady speeds upto 10krpmmore rpm would be better if i can get it to run smoothly enough.im having more problems than I had expected, despite the fact that there is nothing loading the shaft by more than its own weight,Just some optical components at each end, and a BLDC motor in the middle.I dont have much in the way of workshop machinery other than an old pilar drill, and a dremel etc, but id like to improve what ive built so far to at least prove if my experiment is going to work or not before going to the expense of having it made by an engineering firm or something, although most engineering firms I know of wouldnt have much more of a clue than I would.What I realy need is some perfect bearings, but my usual stockist doesnt seem to have any of these,so Im using plain old ball bearings. (single row deep groove)as the shaft goes into whats been described to me as whirl mode at a certain speed, wich is an interesting phenomena wich i think is more complicated than my initial understanding of it, so ive had to use bearings at closer intervals,there a few more difficulties wich im trying to overcome, basically radial play and unequal twisting of the shaft cuased by different friction from the center to each end as both of these are cuasing errors wich are swamping the measurements I am trying to make.the base is made from 2"x4" aluminium rect tube, with 2"x2" alu rect tube for housing the bearings motor and opticals.although this isnt ideal for mounting the bearings, its not the worst of the problems i dont think.the shaft is stainless steel.The ball bearings have a significant amount of play unless pre loaded but preloading increases friction.so is there a better type of bearing ?I was considering needle roller bearings, at least I can get hold of thesebut im not sure how well the radial play is controlled,or how much friction they have.does the fit of the...

C

colin 19 years ago

thanks for the link etc,

I spent some time looking for a supplier that sold bearings that were specifically marked as precision type, the link you gave seemed to suggest that all ball bearings were precision made, wich is kind of true.

however I read about SKF 'explorer' bearings and thought they seemed pretty good, then later while I was looking in my box of bits I saw some of the bearings I had were marked skf explorer, I remember they were quite a bit more expensive, but at the time i thought id see if they were any better, they were just listed the same as all the other bearings.

the old bearings had been through quite a bit and had probably got somewhat dirty anyway, however these new ones are stil just deep groove ball types and still have a considerable amount of play when loose. I looked more for four point contact bearings, but seems to be they are made very large for turntables >>25mm, or very small

Vote

N

Ned Simmons 19 years ago

The internal clearance (what you're describing as "play") is controlled independently of the precision in deep groove bearings. If you look at the complete part number for a specific bearing it will include a C followed by a digit, which signifies the internal clearance. Unless you specify otherwise, the internal clearance in an ABEC7 bearing won't be much different than an ABEC1.

I don't believe all mfrs use exactly the same C codes, but IIRC SKF uses C0 to specify a bearing with minimal clearance and C3 for standard clearance.

Ned Simmons

Vote

C

colin 19 years ago

I realise deep groove have some clearence, these skf explorer have size # = 625-2z as well as 241c on the pcketc and 2067c* on the bearing

Colin =^.^=

Vote

C

colin 19 years ago

Wel I dont quite need 40hp, in fact im using about 10 watts, half of wich is probably wasted in the winding resistance, and theres a fair amount of wind from the motor too, the optical devices dont seem to cuase much wind at all.

I suppose I could use the thing in reverse, blowing air into it so I wouldnt need a motor. my optical devices have 6mm and 2mm dia bore holes in them. and need to be 1meter apart, so I'd need little shafts coming out of the end.

In fact if I was using air to drive it I would consider using air bearings. a compressor might be a bit noisy to leave running overnight in the garage.

although my lashed up prototype is running pretty sweatly upto12krpm atm, I just have to wait a while to see how the errors average out.

There was one bearing in the motor I couldnt replace without some serious surgery. the rest are new, and I think the only noise is coming from the one.

I might toy with some designs for some electromagnetic bearings while im waiting, maybe combine the motor stator windings with it.

oh and im also designing some way to balance it more easily, probably using stroboscope, inductive proximity sensor, and some pairs of slightly eccentric rings on the shaft.

Colin =^.^=

Vote

B

Brian Whatcott 19 years ago

All prop shafts on rear wheel drive cars with stick run at engine speed in top. Faster than the engine in overdrive.

Which reminds me: there was a WWII warplane with a quill shaft under the pilot's seat, connecting the prop and the engine. This had a curved drive line too.

The talk of ABEC 7 bearings reminds me that I bought a little gal a pair of roller skate boots at Christmas.

She wanted a special wheel color on the buckskin boots, and the upgrade came with premium wheel bearings too - ABEC7 if I recall. The bearings are about the size the questioner is using.

Brian Whatcott Altus OK

Vote